KR101595366B1 - Slice-specific phase correction in slice multiplexing - Google Patents

Abstract

The present invention relates to a method for correcting a signal phase at the time of acquiring MR signals to be inspected in a slice multiplexing method, wherein MR signals from at least two different slices of the object to be inspected are simultaneously detected at the time of acquiring the MR signals , The signal phase correction method comprising: determining a linear correction phase in a slice selection direction for each of the at least two slices; Emit an RF excitation pulse having a slice-specific frequency at each of the at least two different slices; Switching slice selection slopes during a slice selection period, wherein different RF excitation pulses are emitted at the at least two different slices during the slice selection period, the slice selection period having a time center in the middle of the slice selection period , Wherein the different RF excitation pulses temporally overlap for the at least two different slices; And determining a time offset of the RF excitation pulse for a time midpoint for each of the RF excitation pulses so that a slice-intrinsic correction slope moment in the slice selection direction corresponding to the linear correction phase of each slice, To the magnetization of the slice of the signal phase.

Description

SLICE-SPECIFIC PHASE CORRECTION IN SLICE MULTIPLEXING < RTI ID = 0.0 >

The present invention relates to a method for correcting a signal phase upon acquiring MR signals to be inspected in a slice multiplexing method, and to an MR system therefor.

In general, the desire for a much faster MR acquisition in a clinical environment is leading to a revival of methods for simultaneously acquiring multiple images. In general, these methods can be characterized in that the transverse magnetization of at least two slices is used at the same time for the imaging process ("multiple slice imaging "," slice multiplexing ") at least during a part of the measurement. Alternatively, in the established "multi-slice imaging ", the signals are obtained alternately from at least two slices, i.e. completely independent of each other, thus increasing the measuring time. For example, among the slice multiplexing methods, there are the following.

Hadamard coding (e.g., Souza et al., J.CAT 12: 1026 (1988)): two (or more) slices are excited simultaneously, Is applied to each slice through a corresponding design of excitation pulses. Signals of magnetization are simultaneously received from both slices. A similar second excitation of both slices is implemented, but has a different relative signal phase at the slices. The remaining imaging process (phase coding steps) is done in a conventional manner; The method may be combined with any acquisition techniques ((multi-) gradient echo, (multi-) spin echo, etc.). The signaling information of both slices can be separated from the two exposures by appropriate computing operations.

Simultaneous echo refocusing (SER, SIR, eg Feinberg et al., MRM 48: 1 (2002)): Two (or more) slices are simultaneously excited. Signals of magnetization are simultaneously received from both slices. During data acquisition, the slope is activated along the slice normals, which leads to the separation of signals of both slices in frequency space. The remaining imaging process (phase coding steps) is done in a conventional manner; The method may be combined with any acquisition techniques ((multi-) gradient echo, (multi-) spin echo, etc.). Images of both slices can be separated from the simultaneous acquisition data by appropriate computing operations.

Wideband data acquisition (eg, Wu et al., Proc. ISMRM 2009: 2768): Two (or more) slices are simultaneously excited. Signals of magnetization are simultaneously received from both slices. During data acquisition, the slope is activated along the slice normals, which leads to the separation of signals of both slices in frequency space. The remaining imaging process (phase coding steps) is done in a conventional manner; The method may be combined with any acquisition techniques ((multi-) gradient echo, (multi-) spin echo, etc.). The signals of both slices can be separated from the simultaneously acquired data by appropriate filtering.

Parallel imaging in the slice direction (e.g., Larkman et al., JMRI 13: 313 (2001)): Two (or more) slices are simultaneously excited. Using at least two (or more) coil elements, the signal of magnetization is simultaneously received from both slices. The remaining imaging process (phase coding steps) is done in a conventional manner; The method may be combined with any acquisition techniques ((multi-) gradient echo, (multi-) spin echo, etc.). Additional calibration measurements are made to determine the spatial acquisition characteristics of the coil elements. The signals of both slices can be separated from the concurrent acquisition data by appropriate computer operations (e.g., the GRAPPA algorithm).

Moreover, in single slice imaging, it may be necessary to correct image artifacts when the correction parameters strongly depend on the spatial location, or the signals of the individual slices, respectively. An example of this is the correction of phase errors that occur due to the accompanying Maxwell field. These phase errors occur because there is no complete linearity of the magnetic field gradient at the switching of the linear magnetic field gradient; Rather, higher order terms always occur. These magnetic fields, called the Maxwell magnetic field, cause phase errors in the detected MR signals. One possibility for correction is described in Meier et al., MRM 60: 128 (2008). In addition, in single slice imaging, it is often necessary to correct for local irregularities in the base field that cause signal erasure or image distortion. Correction of such non-uniformities is described, for example, in Deng et al., MRM 61: 255 (2009) and Lu et al., MRM 62:66 (2009).

In many instances, in the case of slice-intrinsic correction at single-slice exposures, it is sufficient to apply an additional linear signal phase along slice coding. Subsequently, in many instances a sufficient reason for the application of the linear signal phase is explained using various examples.

a) One possibility of applying linear correction is to correct for phase errors in diffusion imaging due to the Maxwell magnetic field.

In MRM 60: 128 (2008), how the accompanying magnetic fields of the Maxwell magnetic fields of the diffusion coding gradients cause additional signal phase deviations along the three spatial coordinate axes. Phase shift along the frequency and phase coding axis only causes displacement of the signal in k space, and the echo is not obtained any more at k = 0, but rather is obtained at the (slightly) shifted position. The echo shift in k space corresponds to the linear phase response in the image (after Fourier transform) in the position space; As far as size images are concerned, this effect only plays a minor role. Furthermore, through the acquisition of a sufficiently large k-space region (e.g. omitting the partial Fourier technique), it can be ensured that in all cases the echo signal is located in the area to be scanned.

However, phase deviations along the slice selection axis directly lead to signal losses that can not be compensated. Thus, the magnitude of the phase deviations depends on the magnitude of the accompanying Maxwell's magnetic field (and thus the position of the slice). For simultaneous acquisition of multiple slices, individual phase deviations must be corrected for all slices. In the first order, the phase shift can be described by a linear phase response.

b) Linear phase correction is sufficient even for correction of phase errors in flow imaging due to the Maxwell magnetic field.

As in the previous example, this is related to the compensation of the phase shift due to the incident Maxwell magnetic field caused by the slopes used in flow coding. The description of the derivation of the linear slice-intrinsic correction phase along the slice coding axis applies here as well. The uncorrected linear phase response in the image (due to the shift of the echo in k space) can be considered in a simple manner in data processing.

c) Linear corrections are also possible (z-shim) to correct for local non-uniformity of the fundamental magnetic field and the signal cancellation caused by them.

MRM 61: 255 (2009) (and in particular references cited in Yang et al., MRM 39: 402 (1998)) discloses imaging in echo plane gradient echo imaging caused by non- It is described how errors can be reduced through repeated implementations of measurements using different auxiliary slopes in the slice coding direction. These are (local) magnetic field gradients that cause phase offsets of signals along three spatial coordinate axes. In addition, only the slope along the slice coding direction has the greatest effect on image quality due to the signal loss (intra-voxel phase shift) in the associated voxel. The method known as the z-shim method changes the background slope per measurement to ensure a good repositioning of each slice in at least one measurement for each of the spatial regions. Multiple images of the slice are merged into the image with reduced signal cancellation either through simple averaging (absolute mean, "sum of squares") or through more complex combination methods.

d) Linear phase correction can also be applied to correction of signal cancellation and image distortion caused by local non-uniformity of the fundamental magnetic field (SEMAC).

MRM 62:66 (2009) describes how signal cancellation and image distortion in 2D imaging due to metal ion implantation (or, respectively, local non-uniformity of the associated fundamental magnetic field) Can be reduced through use. As with the z-key case, a number of measurements using different auxiliary slopes in the slice coding direction are made per slice, and such data are combined in an appropriate manner.

Starting from the above-described prior art, it is now an object of the present invention to provide linear phase correction in a simple manner in a slice multiplexing method in which a specific absorption rate (SAR) and peak RF power of emitted RF pulses are minimized simultaneously .

This object is achieved using the features of the independent claims. Preferred embodiments of the present invention are described in the dependent claims.

According to a first aspect of the present invention, there is provided a method for correcting a signal phase at the time of acquiring MR signals to be inspected in a slice multiplexing method, wherein at the time of acquiring MR signals, at least two different slices Are detected. According to one step of the method, a linear correction phase in the slice selection direction is determined for each of the at least two slices. Moreover, an RF excitation pulse with a slice-natural frequency is emitted at each of at least two different slices simultaneously obtained. Also, the slice selection slope is switched during the slice selection period, and different RF excitation pulses are emitted for at least two different slices during this period. The slice selection period has a time center point at the center of the slice selection period. Moreover, the different RF excitation pulses emitted during the slice selection period overlap for at least two different slices. The time offset of each RF excitation pulse to the time center of the slice selection period is also determined so that the slice-intrinsic correction slope moments in the slice selection direction corresponding to the linear correction phase of each slice are applied to the magnetization of each slice .

Through the time offset of the different RF pulses during the application of the slice selection slope, a readily distinguishable coherency path of the excited spindle is generated for each slice excited by the RF pulses. Are excited by RF pulses applied at earlier times or each refocused signal experiences stronger or different effects than signals from later applied RF pulses due to the additional applied slice selection slopes. Through determination of the time offset, a linear correction phase or a respective linear correction slope moment can be determined for each slice individually, which is necessary to achieve the desired linear correction in each slice. Since the linear phase correction is not the same at all slices and therefore the time offsets of the RF pulses are not the same, a reduction of the specific absorption rate (SAR) is achieved through different time offsets of the RF excitation pulse at the time midpoint within the slice selection period. The SAR is quadratically dependent on the applied pulse voltage and the required peak RF power. If the pulse envelopes for exciting the two slices are emitted simultaneously, they will be superimposed additionally. If both positive extremes are to be in the same position, doubled peak RF amplitudes are required and will therefore produce four times the SAR. Thus, in the case of displacement of the maximum values for the individual slices, the correction slope moment required for each slice can be applied first, and then the SAR can be significantly reduced.

The linear correction phase required for each individual slice may have been previously analytically computed, depending on the application, or may have been determined using previous measurements.

For example, for correction of linear phase errors due to the Maxwell magnetic field, the average magnetic field gradient available for each slice S is initially determined along the slice normal, for example. This can be calculated based on the description in MRM 60: 128 (2008), since the equation for the magnitude of the magnetic field amplitude is developed to the first magnitude using the actual slice position z = z _s .

For example, for correction of linear phase errors due to non-uniformity of the basic magnetic field, the user can specify the area to which the linear magnetic field gradient along the slice normal is to be related. Moreover, the user can specify the resolution to use, i.e. the number of auxiliary slopes to use for each slice. As an alternative, it is possible for the measurement of the non-uniformity of the fundamental magnetic field to be made initially by means of MR methods known to the person skilled in the art. A magnetic field map obtained in such a manner can be evaluated along a slice normal to each slice S with respect to the average magnetic field slopes. The distribution width of this magnetic field gradient can also be determined in such a way. The data obtained in such a manner can be used to individually determine the auxiliary slopes to be used for each slice.

In one embodiment, it is possible to calculate a time offset for each slice that directly depends on the required linear correction phase for each slice, and to apply an RF pulse using the calculated time offset. In another embodiment, it is possible to determine a correction slope moment belonging to each of the different slices, and the average correction slope moment applied to both at least two different slices is determined using a correction slope moment belonging to each slice. Moreover, the deviation of the corresponding slice-inherent correction slope moments is also determined from the average correction slope moments for each of the different slices. In addition, this deviation corresponds to slice-to-individual auxiliary tilt moments, respectively. In addition, a time offset can be calculated for each of the different slices, and thus a slice-individual auxiliary tilt moment acts on each slice. Through the use of the average correction slope moments applied to all slices and the use of slice-individual auxiliary slope moments, the slice-individual auxiliary slope moments are typically less than the correction slope moments pertaining to each slice. Thus, the time offset of the RF excitation pulses with respect to the time center point can be reduced, thus reducing the period as a whole and shortening the acquisition time of the MR signals.

In a further step, it is possible to select the number of slices to be acquired simultaneously according to at least one selection criterion. For example, the best separability of slices in the aliasing process can be used as a selection criterion. This is typically accomplished through a sufficiently large interval of slices. The aliasing process is used to separate the MR signals obtained at the same time of different slices. The other criterion would be an optimal large similarity of the slice correction slope moments needed to require the RF pulses to be shifted inversely to each other as much as possible. A further possibility is to achieve the minimum difference in required slice correction slope moments, so that the minimum spacing of RF pulses is guaranteed for SAR reduction. These different selection criteria may also be arbitrarily combined with each other or may be used separately.

In one embodiment, in addition to the slice-specific correction slope moments, the phase correction along the slice direction due to the Maxwell magnetic field is corrected using the linear correction phase.

In the case of the use of an average correction slope moment, this can be applied in at least two different slices by switching an additional correction slope in the slice direction. For example, this additional correction slope in the slice direction may also be superimposed on the slice re-imagining slope.

As described in the preface of the specification, it is known how a number of different corrective tilt moments are applied on a single slice. According to the present invention, it is possible to apply all the different corrective tilt moments used in different slices to each of at least two slices. For example, if J correction slope moments are applied to each slice (where J > = 2), and N is the number of slices simultaneously acquired, and N is an integer multiple of J, A unique correction slope moment may be applied to each slice and the first step is repeated under permutation of the slice order until J correction slope moments are applied to each slice. However, even more complex permutations are possible when, for example, J is not an integer multiple of N.

In a further embodiment, it is possible for a plurality of RF excitation pulses to be emitted at each slice prior to acquisition of the MR signals, and a plurality of RF excitation pulses are emitted during each slice selection period. A plurality of RF excitation pulses and associated slice selection periods can now be selected with their respective time centers, and thus the slice-intrinsic correction slope moments can be selected as a whole in the magnetization at each slice after switching of multiple RF excitation pulses .

The method can be used not only in the excitation pulses but also in the re-focusing pulses as in, for example, spin echo experiments, or even in storage pulses in stimulated echo experiments, for example. This method can be used for imaging purposes and for excitation of slices for spectroscopic purposes. These refocusing or storage pulses are also applied simultaneously with the slice selection slopes. The invention relates to a module for determining a first linear correction phase or a respective correction slope moment for each of simultaneously excited slices and a module for determining a slice-specific correction slope moment in the slice selection direction corresponding to the determined linear correction phase of each slice To an MR system designed to implement the above method using an MR acquisition sequence controller designed to control the acquisition of MR signals so that they act on the magnetization of each slice.

Hereinafter, the present invention will be described with reference to the accompanying drawings. The following figures are shown.

1 is a schematic diagram of an MR system in which slice-specific correction slope moments can be applied to individual slices in a slice multiplexing method.
2 is a schematic diagram of the time offsets of two RF pulses to excite two different slices upon application of a slice selection slope;
Figure 3 is a flow chart that includes steps when the method according to the present invention is used to correct Maxwell dependent phase errors.
4 is a flow chart including steps when the method is applied when different correction slope moments are repeatedly applied to each slice.
5 is a diagram showing the application of RF pulses and tilt moments according to the prior art.
6 is a diagram illustrating the application of RF pulses in the method according to FIG. 4 according to the present invention.

An MR system in which MR system slice-individual tilt correction moments can be applied to individual slices in a slice multiplexing method according to the present invention is schematically illustrated in FIG. In the present invention, the slopes covered are magnetic field gradients for spatial coding. An MR system 10 having magnets 11 for generating a polarization field B _{0 is} shown in Fig. The examinee 13 disposed on the bed 12 is driven into the MR system. Radio frequency coil arrangements 16 and 17 capable of detecting MR signals from two different slices 14 and 15 are used to detect MR image data from the first slice 14 and the second slice 15 Are schematically shown. The MR system further comprises a tilt system 18 for achieving spatial coding using RF pulses emitted by the RF coils 16,17. As is known, the resulting magnetization in the two slices 14, 15 faces the direction of the fundamental magnetic field before emission of the RF pulses. Through radio frequency coils 16 and 17, radio frequency pulses can be generated, whereby the magnetization in the different slices is deflected from its steady state. Further, two slice MR signals can be detected using the RF coils 16 and 17. The emission of the RF pulses may be done in cooperation with either a whole-body RF coil (not shown) or one or both of the local coils 16, 17. Although individual coils can be used for the transmission of RF pulses (full coil) and for reception of data (local coil), multiple transmit or receive coils may be used.

The manner in which MR signals can in principle be detected through a sequence of magnetic field gradients and emission of RF pulses is known to those skilled in the art and is not described in detail here. Additional modules for controlling the MR system are provided, for example an acquisition controller 20 for controlling the time sequence of the emission of RF pulses and the magnetic field gradients according to the selected imaging sequence. An RF module (21) is provided for controlling the generation of radio frequency pulses in accordance with the control signals of the acquisition controller (20). In addition, a tilt module 22 is provided for controlling the switching of the magnetic field gradients for spatial coding. The operator can control the workflow of MR acquisition through the input unit 23, for example by selecting an appropriate imaging sequence or, in the case of a spectroscopic sequence, an appropriate spectroscopic sequence. In the phase determination module 24, the correction phase to be applied for each single slice in the case of simultaneous acquisition of multiple slices is determined. As mentioned in the preface of the specification, for example, in the correction of diffusion imaging, flow imaging or B ₀ field non-uniformity, it may be desirable to apply a linear signal phase to the excited slice to correct for phase errors. For example, this linear phase to be applied to the slices 14, 15 may be known to the user and may have been input via the input unit 23, and thus the phase determination module accepts the input values. Furthermore, slice-individual phase errors may have been determined in previous measurements and may be stored in the MR system, so that the phase determination module reads the previously stored required phase corrections from the memory. The MR images generated by the MR system 10 may be displayed on the display unit 25. Of course, the MR system has additional components (not shown). However, these have been omitted for clarity, and only the components necessary to understand the present invention are described. Also, it is understood that the different modules and units shown in FIG. 1 are of course designed in different configurations, and need not necessarily be designed as separate units. Different modules or units may be combined differently from one another. Moreover, the different units may be formed through hardware components or through software, or through a combination of hardware and software.

Now, FIG. 2 shows how a somewhat different coherence path occurs for each slice through the selection of a time offset when two RF pulses are given at the time of application of the slice selection slope. For example, two different slices may be slices 14, 15 that are excited simultaneously as shown in FIG. Figure 2 shows this for the case of simultaneous excitation of two slices. In principle, this also applies here to three or more slices. The slice selection slope 26 is switched during the slice selection period 27 and the slice selection period has a time center point represented by a dotted line 28. [ Typically, the slice selection slope also has a negative portion 26a, which is essentially a refraction of the excited spindle, having a length half of the length of the positive portion. An RF excitation pulse that is emitted exactly at the time center - that is, symmetric with respect to the midpoint of time and with a maximum value at the midpoint of time - will produce a slice excitation with complete repositioning of the magnetization along the slice normal. During slice multiplexing, two different RF excitation pulses are emitted and the reading of the MR signals of two slices (not shown) is made simultaneously. An RF excitation pulse 29 having a frequency? 1 and an RF excitation pulse 30 having a frequency? 2, which can represent part of a slice multiplexing method using two shown RF pulses, i.e. two slices obtained at the same time, Are offset by time DELTA T1 or DELTA T2, respectively. As a result, the amplitude correction △ is the slope of the moment T1 G _S applied at the time of the magnetization in the slice excited by the RF pulse (29), wherein G _S is the strength of the slice selection gradient. Further, the tilt moment of the amplitude correction △ T2 G _S applied at the time of the magnetization in the slice excited by the RF pulse (30). Therefore, in the illustrated example,? T2 <0 can be applied. Any correction slope moments can be generated through selection of the corresponding time offset. In this way, in slice multiplexing methods, individual actions of slope along the slice selection axis can be achieved for each of the slices considered simultaneously. This is known as the energy-SAR emitted into the object being inspected-and the required RF peak power can be reduced through each time separation or shift of individual RF excitation pulses. The only switching of the RF pulse to excite one of the slices without crossing with other RF pulses is not necessary so that it can act only on one slice individually. Figure 2 shows the envelopes of two RF pulses 29,30 needed to excite one slice. The resulting RF pulse that simultaneously excites the two slices is generated from the complex addition of the time curves of the amplitude and phase of the two individual pulses. If the positive peak values are located at the same position (DELTA T1 = DELTA T2), then twice the peak RF amplitude is needed and thus four times the SAR will be generated.

3, the following describes how a linear phase change-linear correction phase can be achieved through the application of a single correction slope moment per slice. The method shown in FIG. 3 can be used to correct phase errors that occur in Maxwell magnetic fields, for example, in diffusion imaging or flow imaging.

After the start of the method in step S31, in step S32, the necessary correction slope moments K _i along the slice normal are determined for each slice i. The slice-specific correction slope moment K _i may be known or calculated in advance. Then, in step S33, the number of slices to be simultaneously acquired is selected. For example, the best separability of slices can be used as a selection criterion in the aliasing process, which typically leads to a sufficiently large spacing of slices. It is also possible to use an optimal high similarity of the slice-specific correction slope moments as a criterion so that the RF pulses are shifted opposite each other as much as possible. On the other hand, to ensure minimum spacing of RF pulses for the purpose of SAR reduction, the minimum difference in required slice-specific correction slope moments can be used as an additional basis. Here, "minimum interval" should be understood as a compromise between "predetermined minimum interval &quot;, thus a) sufficiently large separation to reduce the SAR and b) sufficiently small separation to not significantly change the echo time. The above-mentioned criteria for selection of the number of slices can be used individually or in combination.

In step S34, the average correction tilt moment is determined. This average correction slope moment may be applied in common for all slices, for example, through additional individual correction slope moments. This can also be superimposed on the slice reverse slope as the negative slope segment 26a of FIG. However, the separation in time of this common part is advantageous, but not absolutely necessary. In step S35, the deviation from the mean correction tilt moment (M _K ) is determined. This deviation represents the slice-individual auxiliary tilt moments (DELTA M _Ki ) for all simultaneously excited slices, where i is the slice index. Then, in step S36, the required time shift of the individual RF pulses is determined. The RF pulses may be excitation, refocusing or storage RF pulses. If only the RF pulses here are to be shifted, the shift is calculated as ΔT _i = ΔM _Ki / G _S , where ΔM _Ki = M _Ki - M _K. For example, if multiple RF pulses are used, such as in spin echo experiments, each of these pulses may in principle be shifted. Ultimately, only the previously calculated correction slope moment (M _Ki ) should be guaranteed to be applied to the coherence path of magnetization at slice i. For example, partial correction is also possible if the time shift for a complete correction is too large for certain applications. In this case, the defined slice-intrinsic correction phase is not the theoretically preferable phase but the phase that should be achieved in the specific example. Finally, in step S37, the execution of the acquisition of the selected slices is followed, and RF pulses with computed time shifts and calculated cavity correction slope moments are used. In step S38, a check is made as to whether all slices have been acquired. Otherwise, the process continues to step S33, in which a loop is implemented until all desired slices are obtained. The method ends at step S39.

For a given slice thickness S, the amplitude G _S of the slice selection slope may vary within a predetermined limit through the bandwidth BW of the RF pulse. G _S = 2? /? BW / S. Thus, the time separation of the RF pulses can be affected.

A method of applying a number of different correction slope moments to all slices is illustrated in FIG. 4, one of which may be zero. Correction slope moments may be the same for all slices, but this is not an absolute requirement. In the method described in FIG. 4, it is also advantageous to achieve a reduction of SAR and required peak RF power through proper arrangement of RF pulses.

Before describing FIG. 4 in more detail, with reference to FIG. 5, it is again described how different correction tilt moments are applied in the prior art. 5, when switching of the slice selection slope 51 having the positive portion 51a, the negative portion 51b and the portion 51c corresponding to the slice-intrinsic correction slope moment is given, An RF pulse 50a having a resonance frequency? 1 was emitted from the slice. At the same time, during the slice selection period 53 having the time center point 54, the RF pulse 50b having the resonance frequency? 2 with respect to the second slice is emitted simultaneously with the first RF pulse 50a. In the second acquisition, each RF pulse 50a, 50b is emitted with a second correction tilt moment 51d. As is apparent from Fig. 5, the complete overlap of the RF pulse envelopes increases peak RF power and SAR. The disadvantages shown in Fig. 5 can be avoided by the embodiment shown in Fig. The two RF pulses 60a and 60b are time offset with the respective resonant frequencies omega 1 and omega 2 in the first measurement during period 62 while the positive and negative tilt moments 61a and 61b are The slice selection slope 61 is released during the slice selection period. The time offset at time midpoint 63 reaches ΔT1 for pulse 60a and reaches ΔT2 for midpoint 63 for pulse 60b. The application of each of the additional tilt moments 51c or 51d from Figure 5 is not necessary since the slice-specific correction slope moments are generated through the time offset. 6) is assigned to the second RF pulse 60b and the time shift of the second RF pulse 60b is assigned to the RF pulse 60a ). This means that the correction slope moments of each of the simultaneously obtained other slices are applied to each slice of the respective correction slope moments. This method is described in detail again with reference to FIG. According to the state of the method in step S40, the j = 2 ... J correction slope moments (M _Kj ) required for each slice are determined in step S41. As in step S33, in step S42, the number N of slices to be simultaneously acquired is selected. In the simplest example, J is an integer multiple of N, and N is equal to the number of simultaneously obtained slices. In step S42, which is similar to step S34, an average correction slope moment ( _MK ) that can be applied in all measurements is also determined.

In step S44, which is similar to step S35, the deviation of the mean correction slope moments for each slice is determined, where DELTA M _Ki = M _Ki - M _K. This deviation represents the correction tilt moments required for the J measurements. In step S45, which is similar to step S36, the time shift of individual RF pulses which are excitation pulses, refocusing pulses or storage pulses is determined. When only the excitation RF pulses are shifted, the shift acts as ΔT _j = _{ΔM Kj} / G _S. This time shift must be realized for each slice in the measurement, respectively. Otherwise, step S45 is similar to step S36. The execution of the first acquisition of the selected slices with the calculated time shift and the calculated common average correction slope moment (M _K ) is performed in step S46. At step S47, a check is made as to whether all the different correction slope moments of the other slices have been applied to one slice. Otherwise, step S46 is repeated with the replacement of the slice sequence. For example, in the example of N = 3 and J = 3, the following measurements are implemented in steps S46 and S47: Measure 1: Time offset ΔT ₁ is used in slice 1 and time offset ΔT ₂ is used in slice 2 And a time offset? T ₃ is used in slice 3. Additional measurements of the same slice, the time offset △ T ₂ is used in the first slice and the second slice in the time offset △ T _3, a time offset △ T ₁ is used in the third slice. In the third measurement, is applied to the first time slice offset △ T ₃ and time T is applied to the offset △ ₁ to the second slice time offset △ T ₂ is applied to the third slice. Then, in step S48, a check is made as to whether all the slices have been acquired, and steps S42 to S48 are repeated until all slices are obtained.

..., N are used in the first permutation series, J = N + 1, ... 2N is used in the second series, and so forth, if J is an integer multiple of N, then step S46 And S47, a replacement scheme can be adopted in a simple manner.

Of course, more complex permutation schemes may be used. For example, if N = 2 and J = 3 - this means that J is not an integer multiple of N and that a total of four slices are obtained - this can be realized, for example, as follows.

Measure # 1 Slice # 1 ΔT_One Slice # 2 △ T₂

Measurement # 2 Slice # 2 ΔT_One Slice # 1 △ T₃

Measurement # 3 Slice # 1 ΔT₂ Slice # 2 △ T₃

Measurement # 4 Slice # 3? T_One Slice # 4 △ T₂

Measurement # 5 Slice # 4 ΔT_One Slice # 3 △ T₃

Measure # 6 Slice # 3 ΔT₂ Slice # 4 △ T₃

Moreover, in the sequences of more complex permutations, it is possible to consider that a series of required slice-specific correction slope moments is unique to each slice i. For example, the entire set of slices can be divided into P subsets where the same correction slope moments (M _{Kj , p} ) must be applied. The workflow described above in Fig. 4 may also be used for each of these subsets. The method shown in Fig. 4 ends at step S49.

The method described in FIG. 4 may be used not only for correcting undesired phase effects, but also for phase coding itself (known as multi-slab imaging) where a plurality of individually oriented 3D volumes, Lt; / RTI &gt; In this case, the application of the linear phase response is done not only for correcting the undesired signal phase but also for phase coding in 3D imaging. In many individual 3D volumes, multiple excitations with different phase coding slopes in the slice coding direction must be implemented for each of these sub-volumes. This means that in terms of the present invention the linear correction phase is not the correction phase but the phase coding phase and the slice-inherent correction slope moment is not the correction slope moment but the phase coding slope moment applied to each sub-volume . In such a method, the two sub-volumes are simultaneously excited through the time shift of the RF pulses as described above, and a different phase coding slope is applied to each sub-volume. The required phase coding step is then implemented with appropriate permutations for each subvolume, as described in connection with FIG. 4 for different correction slope moments. In such a method, the linear phase coding phase in the slice selection direction will be determined for each slice or for each subvolume in one step, and the RF pulses are determined by the slice-unique phase coding slope moment in the slice selection Direction, and the slice-unique phase coded slope moments correspond to specific linear phase coded phases of each slice. The phase coding slopes along the slice normal required in this method are determined from the acquisition parameters, such as the magnitude of the subvolume in the direction of the slice normal and the resolution in this direction, or from the resolution in this direction, Area. &Lt; / RTI &gt;

In short, the present invention enables slice-intrinsic correction of image artifacts in slice multiplexing methods given simultaneous reduction of SAR and peak RF power.

Claims (14)

A method for correcting a signal phase at the time of acquiring MR signals to be inspected in a slice multiplexing method,
MR signals from at least two different slices (14, 15) of the object to be inspected (13) are simultaneously detected at the time of acquisition of the MR signals,
The signal phase correction method
Determining a linear correction phase in the slice selection direction for each of the at least two slices (14, 15);
Emit an RF excitation pulse (29, 30, 60a, 60b) having a slice-specific frequency at each of the at least two different slices,
(26, 61) during a slice selection period (27), wherein the different RF excitation pulses (29, 30, 60a) at the at least two different slices And 60b are emitted and the slice selection period 27 has a time center point 28 in the middle of the slice selection period and the different RF excitation pulses 29,30,60a, Temporally overlap for different slices; And
Determine a time offset of the RF excitation pulse for a time midpoint (27) for each of the RF excitation pulses (29,30, 60a, 60b) to determine a time offset of the RF excitation pulse for each of the slice selection directions In which the slice-intrinsic corrective tilt moments at the respective slices (14, 15)
/ RTI &gt; The method according to claim 1,
Wherein the correction slope moments belonging to the slice are determined for each of the at least two different slices (14, 15), and an average correction slope moment applied to all of the at least two different slices is determined for each of the slices Wherein a deviation of a corresponding slice-specific correction slope moment with respect to said average correction slope moment is determined for each of said at least two different slices (14, 15), said deviation being determined using slice- Characterized in that for each of said at least two different slices (14, 15), said time offset is calculated such that said slice-individual auxiliary tilt moments act on said respective slice Phase correction method. 3. The method according to claim 1 or 2,
Wherein the number of slices to be obtained simultaneously is further selected according to at least one selection criterion. 3. The method according to claim 1 or 2,
Wherein said correcting slope moment having a linear correction phase corrects phase deviations along a slice direction due to Maxwell fields. 3. The method of claim 2,
Wherein the average correction slope moment is impressed on the at least two different slices (14, 15) through switching of an additional correction slope in the slice direction. 3. The method according to claim 1 or 2,
A plurality of RF excitation pulses (29, 30, 60a, 60b) are emitted from one slice before acquisition of the MR signals and the plurality of RF excitation pulses (29, 30, 60a, 60b) And wherein the plurality of RF excitation pulses (29, 30, 60a, 60b) and associated slice selection periods are selected such that the slice-intrinsic correction slope moment is applied to the magnetization at each slice after switching of the plurality of RF excitation pulses Is selected to be applied as a whole. 3. The method according to claim 1 or 2,
Characterized in that a single slice-intrinsic correction slope moment is applied to each of said at least two different slices (14, 15). 3. The method according to claim 1 or 2,
Wherein a plurality of different corrective tilt moments are applied to each of said at least two different slices (14,15), and all different corrective tilt moments used in said at least two slices are applied to said at least two slices (14,15) Respectively. &Lt; / RTI &gt; 9. The method of claim 8,
J correction slope moments are applied to each slice (where J &gt; = 2), N is the number of simultaneously obtained slices, N is an integer multiple of J, and the slice-intrinsic correction slope moments Slice, and said first step is repeated with permutation of the slice order until said J corrective tilt moments are applied to each slice. 3. The method according to claim 1 or 2,
Wherein the RF excitation pulses (29, 30, 60a, 60b) are excitation pulses for selection of a slice from the object to be inspected. 3. The method according to claim 1 or 2,
Wherein the RF excitation pulses are refocusing pulses at the time of generating a spin echo signal. 3. The method according to claim 1 or 2,
Wherein the RF excitation pulses are storage pulses during signal generation by stimulated echoes. An MR system (10) designed to correct a signal phase upon acquisition of MR signals to be inspected in a slice multiplexing method,
MR signals from at least two different slices (14, 15) to be inspected are simultaneously detected at the time of acquisition of the MR signals,
The MR system (10)
A module (24) for determining a first linear correction phase in a slice selection direction for each of said at least two slices (14, 15); And
Is designed to emit an RF excitation pulse (29, 30, 60a, 60b) having a slice-specific frequency at each of the at least two different slices and is designed to switch slice selection slopes (26, 61) during a slice selection period - the different RF excitation pulses (29, 30, 60a, 60b) are emitted in the at least two different slices during the slice selection period and the slice selection period (27) Wherein the RF excitation pulses (29, 30, 60a, 60b) are temporally overlapping with respect to the at least two different slices, and wherein each of the RF excitation pulses Determining a time offset of the RF excitation pulse with respect to the midpoint and determining a time-offset of the slice-intrinsic correction in the slice selection direction corresponding to the linear correction phase of each slice, An MR acquisition sequence controller 20 designed such that a positive slope moment acts on the magnetization of each slice,
&Lt; / RTI &gt; 14. The method of claim 13,
Wherein the MR acquisition sequence controller is designed to implement the signal phase correction method according to claim 2.

Images